s/floating point/floating-point/ when used attributively.

man2/prctl.2

@@ -153,11 +153,11 @@ in the location pointed to by
 .BR PR_SET_FPEXC " (since Linux 2.4.21, 2.5.32, only on PowerPC)"
 Set floating-point exception mode to \fIarg2\fP.
 Pass \fBPR_FP_EXC_SW_ENABLE\fP to use FPEXC for FP exception enables,
-\fBPR_FP_EXC_DIV\fP for floating point divide by zero,
-\fBPR_FP_EXC_OVF\fP for floating point overflow,
-\fBPR_FP_EXC_UND\fP for floating point underflow,
-\fBPR_FP_EXC_RES\fP for floating point inexact result,
-\fBPR_FP_EXC_INV\fP for floating point invalid operation,
+\fBPR_FP_EXC_DIV\fP for floating-point divide by zero,
+\fBPR_FP_EXC_OVF\fP for floating-point overflow,
+\fBPR_FP_EXC_UND\fP for floating-point underflow,
+\fBPR_FP_EXC_RES\fP for floating-point inexact result,
+\fBPR_FP_EXC_INV\fP for floating-point invalid operation,
 \fBPR_FP_EXC_DISABLED\fP for FP exceptions disabled,
 \fBPR_FP_EXC_NONRECOV\fP for async non-recoverable exception mode,
 \fBPR_FP_EXC_ASYNC\fP for async recoverable exception mode,

man2/sigaction.2

@@ -406,19 +406,19 @@ integer divide by zero
 integer overflow
 .TP
 .B FPE_FLTDIV
-floating point divide by zero
+floating-point divide by zero
 .TP
 .B FPE_FLTOVF
-floating point overflow
+floating-point overflow
 .TP
 .B FPE_FLTUND
-floating point underflow
+floating-point underflow
 .TP
 .B FPE_FLTRES
-floating point inexact result
+floating-point inexact result
 .TP
 .B FPE_FLTINV
-floating point invalid operation
+floating-point invalid operation
 .TP
 .B FPE_FLTSUB
 subscript out of range

man3/INFINITY.3

@@ -22,7 +22,7 @@
 .\"
 .TH INFINITY 3  2007-07-26 "" "Linux Programmer's Manual"
 .SH NAME
-INFINITY, NAN, HUGE_VAL, HUGE_VALF, HUGE_VALL \- floating point constants
+INFINITY, NAN, HUGE_VAL, HUGE_VALF, HUGE_VALL \- floating-point constants
 .SH SYNOPSIS
 .nf
 .B #define _ISOC99_SOURCE

man3/__setfpucw.3

@@ -13,9 +13,9 @@ __setfpucw \- set FPU control word on i386 architecture (obsolete)
 .BR __setfpucw ()
 transfers
 .I control_word
-to the registers of the FPU (floating point unit) on the i386 architecture.
-This was used to control floating point precision,
-rounding and floating point exceptions.
+to the registers of the FPU (floating-point unit) on the i386 architecture.
+This was used to control floating-point precision,
+rounding and floating-point exceptions.
 .SH CONFORMING TO
 This function was a non-standard GNU extension.
 .SH NOTES
@@ -25,7 +25,7 @@ There are new functions from C99, with prototypes in
 to control FPU rounding modes, like
 .IR fegetround ,
 .IR fesetround ,
-and the floating point environment, like
+and the floating-point environment, like
 .IR fegetenv ,
 .IR feholdexcept ,
 .IR fesetenv ,

man3/fenv.3

@@ -28,7 +28,7 @@
 feclearexcept, fegetexceptflag, feraiseexcept, fesetexceptflag,
 fetestexcept, fegetenv, fegetround, feholdexcept, fesetround,
 fesetenv, feupdateenv, feenableexcept, fedisableexcept,
-fegetexcept \- floating point rounding and exception handling
+fegetexcept \- floating-point rounding and exception handling
 .SH SYNOPSIS
 .nf
 .B #include <fenv.h>
@@ -59,18 +59,18 @@ fegetexcept \- floating point rounding and exception handling
 Link with \fI\-lm\fP.
 .SH DESCRIPTION
 These eleven functions were defined in C99, and describe the handling
-of floating point rounding and exceptions (overflow, zero-divide etc.).
+of floating-point rounding and exceptions (overflow, zero-divide etc.).
 .SS Exceptions
 The DivideByZero exception occurs when an operation on finite numbers
 produces infinity as exact answer.
 .LP
 The Overflow exception occurs when a result has to be represented as a
-floating point number, but has (much) larger absolute value than the
-largest (finite) floating point number that is representable.
+floating-point number, but has (much) larger absolute value than the
+largest (finite) floating-point number that is representable.
 .LP
 The Underflow exception occurs when a result has to be represented as a
-floating point number, but has smaller absolute value than the smallest
-positive normalized floating point number (and would lose much accuracy
+floating-point number, but has smaller absolute value than the smallest
+positive normalized floating-point number (and would lose much accuracy
 when represented as a denormalized number).
 .LP
 The Inexact exception occurs when the rounded result of an operation
@@ -156,7 +156,7 @@ The
 function sets the rounding mode as specified by its argument
 and returns zero when it was successful.
 .SS "Floating point environment"
-The entire floating point environment, including
+The entire floating-point environment, including
 control modes and status flags, can be handled
 as one opaque object, of type
 .IR fenv_t .
@@ -170,7 +170,7 @@ ISO C to have round to nearest, all exceptions cleared and a non-stop
 .LP
 The
 .BR fegetenv ()
-function saves the current floating point environment in the object
+function saves the current floating-point environment in the object
 .IR *envp .
 .LP
 The
@@ -182,7 +182,7 @@ It returns zero when successful.
 .LP
 The
 .BR fesetenv ()
-function restores the floating point environment from
+function restores the floating-point environment from
 the object
 .IR *envp .
 This object must be known to be valid, for example, the result of a call to
@@ -224,12 +224,12 @@ It is only defined if
 .B _GNU_SOURCE
 is defined.
 The C99 standard does not define a way to set individual bits in the
-floating point mask, for example, to trap on specific flags.
+floating-point mask, for example, to trap on specific flags.
 glibc 2.2 supports the functions
 .BR feenableexcept ()
 and
 .BR fedisableexcept ()
-to set individual floating point traps, and
+to set individual floating-point traps, and
 .BR fegetexcept ()
 to query the state.
 .sp

man3/finite.3

@@ -23,7 +23,7 @@
 .TH FINITE 3  2007-07-26 "" "Linux Programmer's Manual"
 .SH NAME
 finite, finitef, finitel, isinf, isinff, isinfl, isnan, isnanf, isnanl \-
-BSD floating point classification functions
+BSD floating-point classification functions
 .SH SYNOPSIS
 .nf
 .B #include <math.h>

man3/printf.3

@@ -934,7 +934,7 @@ comes from untrusted user input, it may contain \fB%n\fP, causing the
 .BR printf ()
 call to write to memory and creating a security hole.
 .\" .PP
-.\" Some floating point conversions under early libc4
+.\" Some floating-point conversions under early libc4
 .\" caused memory leaks.
 .SH EXAMPLE
 .if \w'\*(Pi'=0 .ds Pi pi

man3/scanf.3

@@ -683,7 +683,7 @@ or
 .B _GNU_SOURCE
 is also specified), in which case the
 .B a
-is interpreted as a specifier for floating point numbers (see above).
+is interpreted as a specifier for floating-point numbers (see above).
 
 Since version 2.7, glibc also provides the
 .B m

man3/significand.3

@@ -4,7 +4,7 @@
 .TH SIGNIFICAND 3 2007-07-26 "GNU" "Linux Programmer's Manual"
 .SH NAME
 significand, significandf, significandl \-
-get mantissa of floating point number
+get mantissa of floating-point number
 .SH SYNOPSIS
 .B #include <math.h>
 .sp

man3/strtod.3

@@ -42,7 +42,7 @@
 .\"
 .TH STRTOD 3 2007-07-26 "Linux" "Linux Programmer's Manual"
 .SH NAME
-strtod, strtof, strtold \- convert ASCII string to floating point number
+strtod, strtof, strtold \- convert ASCII string to floating-point number
 .SH SYNOPSIS
 .B #include <stdlib.h>
 .sp

man5/acct.5

@@ -84,7 +84,7 @@ enum {          /* Bits that may be set in ac_flag field */
 .PP
 The
 .I comp_t
-data type is a floating point value consisting of a 3-bit, base-8 exponent,
+data type is a floating-point value consisting of a 3-bit, base-8 exponent,
 and a 13-bit mantissa.
 A value,
 .IR c ,

man7/bootparam.7

@@ -149,7 +149,7 @@ is defined.)
 Some i387 coprocessor chips have bugs that show up when used in 32 bit
 protected mode.
 For example, some of the early ULSI-387 chips would
-cause solid lockups while performing floating point calculations.
+cause solid lockups while performing floating-point calculations.
 Using the 'no387' boot arg causes Linux to ignore the maths
 coprocessor even if you have one.
 Of course you must then have your